RUNX1 and CBFB are not only important for leukemogenesis but they are also key regulators of normal hematopoiesis. These two genes are required during the earliest steps of hematopoietic stem cell formation and in many subsequent stages of many blood lineages. Multiple studies suggest that dysregulation of the normal transcriptional program controlled by RUNX1 and CBFB is likely to be an important mechanism for leukemogenesis. Therefore, better understanding of the RUNX1/CBFB transcriptional program and the roles of RUNX1 and CBFB in normal hematopoiesis will lead to better understanding of the mechanisms for leukemogenesis. We have been pursuing two specific aims in this project in the last fiscal year. In the first specific aim, we have been determining the roles of CBFB and RUNX1 in the formation of hematopoietic stem cells (HSCs) in zebrafish. In particular, we tried to discover if HSCs could form in the absence of RUNX1. Previous studies suggest that RUNX1 is required for the emergence of definitive hematopoietic stem cells (HSCs) from the hemogenic endothelium during embryo development. For example, Runx1 knockout mouse embryos lack all definitive blood lineages and cannot survive past embryonic day 13. However, we recently showed that zebrafish homozygous for an ENU-induced nonsense mutation in runx1 (runx1W84X/W84X) were able to recover from a larval bloodless phase and develop to fertile adults with multi-lineage hematopoiesis, suggesting the formation of runx1-independent adult HSCs. However, our finding was based on a single zebrafish mutant line, which requires verification in independent mutants. In order to further investigate if a RUNX1-independent pathway exists for the formation of adult HSCs, we generated two new runx1 mutants using the transcription activator-like effector nuclease (TALEN) technology. These mutations cause frameshifts and premature terminations, resulting in loss of function of runx1 (runx1-/-). Similar to previously described RUNX1 knockout mutants, these new runx1-/- mutant embryos failed to develop definitive hematopoiesis. Time-lapse recordings with confocal microscopy confirmed the absence of HSC emergence in the runx1-/- embryos. The runx1-/- larvae lost circulating primitive blood cells and became bloodless between 8 and 14 days post fertilization (dpf). However they gradually regained circulating blood cells one week later. Eventually, about 40% of runx1-/- mutants developed to fertile adults with circulating blood cells of multi-lineages. Taken together, our data is consistent with the previously described runx1W84X/W84X phenotype and supports the existence of a runx1-independent mechanism for HSC formation and definitive hematopoiesis. In the next fiscal year we will conduct research to discover the pathway for the RUNX1-independent HSC formation. In the second aim we have been using human induced pluripotent stem cells (iPSCs) to study the function of RUNX1 in human hematopoiesis. We have generated iPSCs from patients with familial platelet disorder, an autosomal dominant disease with reduced platelet numbers and increased susceptibility to leukemia. We have also been working on the culturing conditions for the differentiation of iPSCs to hematopoietic cells, in collaboration with scientists at the National Center for Advancing Translational Sciences. Finally, we have been using genomic technology to determine the genomic integrity of iPSCs. Genomic integrity of iPSCs has been extensively studied in recent years and one central question that still remains is the origin of genomic variations in iPSCs. Previous studies have reported conflicting results regarding when genomic variations occur in iPSCs, whether they are inherited from the parental somatic cells, or generated during the reprogramming process. We have undertaken a unique approach to determine the history of genomic variations by deriving single cell clonal fibroblast lines and clonal iPSC lines from the same parental fibroblasts and applied next generation sequencing technologies to detect the DNA sequence variations. Moreover, we interrogated larger DNA alterations by SNP arrays to identify structural variations in each samples. Whole exome sequencing (WES) followed by targeted deep re-sequencing revealed that most single nucleotide variants and small indels that were initially identified as de novo variants in the daughter cells from WES were rare pre-existing variants with very low allele frequency. Deep re-sequencing showed that only less than 0.1% of total variants discovered from WES were completely absent in the parental fibroblasts. Copy number variations (CNVs) tend to have emerged de novo in the iPSCs and fibroblast subclones than inherited from the parental fibroblasts, but each iPSC/fibroblast subclone carried only 1-3 CNVs. Overall our data showed that more than 99.9% of genomic variations present in iPSCs are inherited from parental somatic cells rather than as consequences of reprogramming.